scraper — Claude Code MCP server: features and setup
Pulls readable page text, links, and metadata from URLs. Use it when an AI assistant needs web documentation or article content without manual copying.
claude mcp add scraper -- npx -y mcp-server-scraperWhat scraper does
The server fetches a URL and provides separate tools for page text, links, and metadata. It can also find matching lines on a page or process several URLs, returning a title and excerpt for each one.
It parses HTML in Node with Mozilla Readability and linkedom, without launching a headless browser. No API key is needed. Pages with content already present in their HTML are the best fit; client-rendered SPAs are outside its reach, so those require a browser MCP server.
What it helps with
- turn a documentation page into Markdown
- collect page links and their anchor text
- retrieve a URL’s title, description, and OG tags
- find matching lines within a web page
- process several URLs and get an excerpt for each
Example prompts for the agent
- Read
https://docs.example.com, extract its content, and summarize the page. - Collect every link on this page with its anchor text.
- Search this page for
authenticationand retrieve the URL’s title and description.
How to install scraper
Platform CLI
claude mcp add scraper -- npx -y mcp-server-scraperPlatform CLI
codex mcp add scraper -- npx -y mcp-server-scraperConfig file
{
"mcpServers": {
"scraper": {
"command": "npx",
"args": [
"-y",
"mcp-server-scraper"
]
}
}
}Config file
{
"servers": {
"scraper": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"mcp-server-scraper"
]
}
}
}Platform CLI
gemini mcp add scraper npx -y mcp-server-scraperStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about scraper
How do I install scraper in Claude Code?
Run claude mcp add scraper -- npx -y mcp-server-scraper. It starts through npx and does not require an API key.
What does scraper do?
This MCP server extracts readable page text as Markdown, along with links and metadata. It can also search page content and handle multiple URLs.
Can scraper handle JavaScript-rendered sites?
No. It parses fetched HTML, so pages whose content is rendered client-side are not a fit. Use a browser MCP server for those pages.
